RESEARCH ARTICLE Open Access

Isolation, purification and characterizationof an ascorbate peroxidase from celery andoverexpression of the AgAPX1 geneenhanced ascorbate content and droughttolerance in ArabidopsisJie-Xia Liu, Kai Feng, Ao-Qi Duan, Hui Li, Qing-Qing Yang, Zhi-Sheng Xu and Ai-Sheng Xiong*

Abstract

Background: Celery is a widely cultivated vegetable abundant in ascorbate (AsA), a natural plant antioxidantcapable of scavenging free radicals generated by abiotic stress in plants. Ascorbate peroxidase (APX) is a plantantioxidant enzyme that is important in the synthesis of AsA and scavenging of excess hydrogen peroxide.However, the characteristics and functions of APX in celery remain unclear to date.

Results: In this study, a gene encoding APX was cloned from celery and named AgAPX1. The transcription level ofthe AgAPX1 gene was significantly upregulated under drought stress. AgAPX1 was expressed in Escherichia coli BL21(DE3) and purified. The predicted molecular mass of rAgAPX1 was 33.16 kDa, which was verified by SDS-PAGE assay.The optimum pH and temperature for rAgAPX1 were 7.0 and 55 °C, respectively. Transgenic Arabidopsis hosting theAgAPX1 gene showed elevated AsA content, antioxidant capacity and drought resistance. Less decrease in netphotosynthetic rate, chlorophyll content, and relative water content contributed to the high survival rate oftransgenic Arabidopsis lines after drought.

Conclusions: The characteristics of APX in celery were different from that in other species. The enhanced droughtresistance of overexpressing AgAPX1 in Arabidopsis may be achieved by increasing the accumulation of AsA,enhancing the activities of various antioxidant enzymes, and promoting stomatal closure. Our work provides newevidence to understand APX and its response mechanisms to drought stress in celery.

Keywords: Ascorbate peroxidase, Purification, Overexpression, Arabidopsis, Drought stress, Celery

BackgroundUnder abiotic stresses, plants generate numerous reactiveoxygen species (ROS), including superoxide radical, hydro-gen peroxide (H2O2), and lipid peroxides. Excessive accu-mulation of ROS damages plant cells via lipid peroxidationand protein oxidation [1, 2]. Many enzymatic and non-enzymatic antioxidants can reduce ROS levels to maintaincellular redox balance [1]. Plants contain various enzymatic

antioxidants, such as superoxide dismutase (SOD),peroxidase (POD), catalase (CAT), and ascorbate per-oxidase (APX), as well as non-enzymatic antioxidants,such as ascorbate (AsA), glutathione, flavonoids andcarotenoids for ROS scavenging system [3, 4].APX of the heme peroxidase superfamily [5, 6] is

involved in the recycling pathway of AsA and environ-mental stress response in plants. AsA is a well-knownplant antioxidant. Natural antioxidants have attractedincreasing attention because of their beneficial effects onhumans [7]. Plant resistance is also an enduring topic inthe field of plant science [8, 9]. Several researchersfocused on improving the ability of plants to tolerate

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: xiongaisheng@njau.edu.cnState Key Laboratory of Crop Genetics and Germplasm Enhancement,Ministry of Agriculture and Rural Affairs Key Laboratory of Biology andGermplasm Enhancement of Horticultural Crops in East China, College ofHorticulture, Nanjing Agricultural University, 1 Weigang, Nanjing 210095,China

Liu et al. BMC Plant Biology (2019) 19:488 https://doi.org/10.1186/s12870-019-2095-1

http://crossmark.crossref.org/dialog/?doi=10.1186/s12870-019-2095-1&domain=pdfhttp://orcid.org/0000-0002-7900-5001http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/publicdomain/zero/1.0/mailto:xiongaisheng@njau.edu.cn

abiotic stresses, including saline-alkali [10], drought [11],heat [12], and cold [13, 14]. APX is a key enzyme in theAsA–glutathione (AsA–GSH) pathway that removesexcessive H2O2 in plant cells under normal and stressconditions [15]. APX catalyzes the reduction of H2O2using AsA as an electron donor to water and monodehy-droascorbate [16]. Then monodehydroascorbate spon-taneously transforms into dehydroascorbate.APX is widely studied because of its important roles in

higher plants. It is also found in eukaryotic algae [17]. APXhas been purified and studied in tea [18], rice [19], komat-suna [20], potato [21], and many other species. The physic-chemical properties of APXs vary in different species.Drought is still the most vital constraint in crop pro-

duction and food security, especially in areas where agri-cultural water resources are insufficient. Global warmingand increasing droughts have aggravated the loss of theagricultural economy [22]. Many studies have reportedthat APX is involved in AsA accumulation and abioticstresses such as salt stress [23–25]. However, relatedreports in resisting drought stress are poor.Celery is a vegetable belonging to Apiaceae with rich

nutritional value, and its growth is influenced by multipleabiotic stresses [26–28]. In our previous study, we deter-mined changes in the expression of the APX gene duringthe developmental stages of celery [29]. However, thecharacteristics and transcriptional regulation mechanism ofAPX under drought stress in celery remain unknown. Inthe present study, the AgAPX1 gene was cloned fromcelery, and its expression levels under simulated droughtstress were detected. The active recombinant proteinrAgAPX1 was obtained and then characterized. Fur-thermore, the subcellular localization of AgAPX1 wasinvestigated. The AsA content, antioxidant capacity,and drought tolerance of Arabidopsis overexpressingAgAPX1 were also detected and analyzed. The resultsof this study imply that AgAPX1 can enhance thedrought tolerance and serve as a potential target forresistance breeding with gene engineering in celery.

ResultsNucleotide sequence and deduced amino acid sequenceof AgAPX1As shown in Additional file 1, AgAPX1 cDNA consists of753 bp nucleotides and encodes 250 amino acids. Themolecular mass and theoretical pI of native AgAPX1 were27.72 kDa and 5.41, respectively. The protein formula wasC1244H1926N332O371S8 with a weak hydrophilicity. Thededuced amino acid sequence of AgAPX1 was alignedwith homologous sequences from other species, includingPisum sativum, Arabidopsis thaliana, Brassica rapa, andOryza sativa using the ESPript 3.0 website [30]. The simi-larity of APX sequences between celery and P. sativumwas the highest (80.8%) (Fig. 1a). Phylogenic tree showed

that the APX proteins from Apium graveolens andSpuriopimpinella brachycarpa belonged to the samebranch, and the APX proteins from other plants in samefamily were also clustered together (Fig. 1b). AgAPX1 andthe APX proteins from other monocot plants were distantin evolution. In addition, the AgAPX1 protein has a spe-cific peroxidase domain from sites 25 to 227 (Fig. 1c).Three-dimensional structure analysis using the pea pro-tein (PDB ID:1apx) as template indicated that AgAPX1contains thirteen TM α-helices and two β-sheets (Fig. 1d).

Expression profiles of AgAPX1 under PEG 6000 treatmentin celeryThe expression profiles of AgAPX1 were detected toverify its response under dehydration stress in celery. Asshown in Fig. 2, the expression level of the AgAPX1 geneat 1 h under PEG treatment was 2.56 folds higher thanthat of the control, and peaked at 4 h followed by adecrease. The results indicated that the AgAPX1 generesponsed to PEG 6000 treatment and involved indehydration stress in celery.

Expression of AgAPX1 in Escherichia coli and purificationof recombinant AgAPX1The AgAPX1 gene was cloned into the pET30 vector to con-struct a protein expression vector (pET-30a(+)-AgAPX1)and then expressed in E. coli BL21(DE3). A purified re-combinant protein of AgAPX1 was prepared andnamed rAgAPX1. Its molecular weight was 33.16 kDaas calculated by ExPASy. Coomassie-stained SDS-PAGE of the purified rAgAPX1 protein showed a singleband at about 34 kDa, which corresponds to the calcu-lation (Additional file 2).

Characterization of rAgAPX1The optimum pH for the rAgAPX1-catalyzed oxidation ofAsA was pH 7.0 (Fig. 3a). The relative enzyme activity ofrAgAPX1 was low in the acid environment, maintained arising activity until pH 7.0, and then decreased. The resultindicates that the suitable reaction environment of the en-zyme is neutral. The optimum temperature for rAgAPX1was found to be 55 °C (Fig. 3b). The enzyme activity ofrAgAPX1 increased with reaction temperature from 20 °Cto 55 °C, peaked at 55 °C followed by a decrease, and thendeactivated at 90 °C.

Subcellular localization of the AgAPX1 proteinThe empty vector (pA7-GFP) and recombinant vector(AgAPX1-GFP) were expressed in onion epidermal cellsto investigate the subcellular localization of AgAPX1.The onion epidermal cell with empty vector displayedstrong fluorescence throughout the entire cell (Fig. 4a).Meanwhile, the onion epidermal cells transformed bythe recombinant vector showed a similar distribution of

Liu et al. BMC Plant Biology (2019) 19:488 Page 2 of 13

green fluorescence to the empty vector (Fig. 4b). The35S:AgAPX1-GFP fusion-construct was also expressedin Arabidopsis mesophyll protoplasts. The green fluores-cence signals were mainly displayed in the nucleus andthe plasma membrane, and did not overlap with chloro-plast fluorescence (Fig. 4c). These results suggest thatAgAPX1 is located in the nucleus and membrane andnot in the chloroplast.

Heterologous expression of AgAPX1 in Arabidopsisincreased the AsA content and total antioxidant capacityTransgenic Arabidopsis plants overexpressing the AgAPX1gene were generated to further investigate the molecular

functions of AgAPX1. The positive transgenic lines hostingthe AgAPX1 gene were confirmed by gene-specific ampli-fication (Fig. 5a). Two transgenic lines showed blue by β-glucuronidase (GUS) staining (Fig. 5b). AsA levels wereassessed using 4-week-old Arabidopsis leaves by HPLC.The AsA contents in the two transgenic lines were mark-edly higher than that in the wild-type (WT) Arabidopsisplants (Fig. 5c and Additional file 3). The AgAPX1–4 linehad the highest level of AsA, which was approximately 1.4times higher than that in the WT plants. The total antioxi-dant capacities of the AgAPX1–4 and AgAPX1–16 linesincreased by 29.10 and 21.16% as evaluated by FRAPassays, respectively (Fig. 5d).

Fig. 1 Sequence characteristic analysis AgAPX1 protein. a Alignment of deduced amino acid sequences of AgAPX1 with other ascorbate peroxidasefrom Pisum sativum (Accession No. P48534.2), Arabidopsis thaliana (Accession No. Q05431.2), Brassica rapa (Accession ACV92696.1), and Oryza sativa(Accession No. Q10N21.1) contained secondary structure elements (helices with squiggles, β-strands with arrows and turns with TT letters). bPhylogenetic relationships of deduced amino acid sequences of AgAPX1 and other APX proteins. Arabidopsis thaliana (Q05431.2), Brassica rapa(ACV92696.1), Camellia azalea (AKP06507.1), Ipomoea batatas (ALP06091.1), Ipomoea trifida (AGT80152.1), Nicotiana tabacum (AAA86689.1), Oryza sativa(Q10N21.1), Pisum sativum (P48534.2), Spuriopimpinella brachycarpa (AAF22246.1), and Zea mays (NP_001105500.2). c The predicted domain location ofAgAPX1. d Three-dimensional structures of AgAPX1. The red and blue parts mark the alpha helix and beta turn, respectively

Liu et al. BMC Plant Biology (2019) 19:488 Page 3 of 13

Overexpression of AgAPX1 in Arabidopsis positivelyregulates drought tolerance by regulating the stomataapertureTransgenic and WT Arabidopsis plants were treated with400mM mannitol to study the tolerance to drought stress.The phenotype of Arabidopsis plants was observed after 7days of mannitol treatment (Fig. 6a). The transgenic linesshowed a better status under drought condition. The sto-matal apertures of WT and transgenic lines both decreasedin response to stress. The two transgenic Arabidopsis linesshowed smaller stomatal apertures after 24 h of droughttreatment. The width-to-length ratio of stomatal aperturein the two transgenic Arabidopsis plants decreased to 0.85and 0.66 times as much as that in the WT plants (Fig. 6b).

Analysis of antioxidant enzyme activities in transgeniclines before and after treatmentThe activities of four key enzymes, APX, SOD, POD, andCAT, were measured in WT and two transgenic Arabidop-sis lines under normal condition and drought treatment.APX, SOD, and POD are major enzymes in the ROS sca-venging system. CAT is also an important antioxidant en-zyme in plants [31]. In plant life cycle, H2O2 accumulatesunder various abiotic stresses [1]. Excessive H2O2 oxidizesbiological macro-molecules (nucleic acids and proteins)directly or indirectly, and damages cell membranes, thusaccelerating cell senescence and disintegration. CAT canscavenge H2O2 and maintain the balance of ROS meta-bolism in plants. The activity of CAT in plant tissues is alsoclosely related to plant stress resistance [31, 32].The transgenic Arabidopsis lines had significantly

higher APX activities than the WT plants (Fig. 6c). TheAPX activities of all samples increased after droughttreatment, and the increase was more remarkable in theAgAPX1–4 transgenic line. In addition, drought stressenhanced the SOD activities in the two transgenic linescompared with the WT plants (Fig. 6d). The POD acti-vities in all of the tested samples increased after droughtstress (Fig. 6e). However, the CAT activities in the WTplants showed markedly increased than those in the twotransgenic lines after drought treatment (Fig. 6f).

Physiological changes in Arabidopsis leaves exposed todrought stressThe net photosynthetic rate (Pn) was measured to eva-luate the photosynthetic activity in WT and transgenicArabidopsis under drought stress. As shown in Fig. 7a,Pn considerably decreased in the WT over transgeniclines under drought stress. The results suggested thatthe photosynthetic system of WT incurred more severe

Fig. 2 Expression profile of AgAPX1 gene under drought stress in‘Jinnanshiqin’. Different letters represent significant difference at 0.05 level

Fig. 3 Effect of pH (a) and temperature (b) on purified AgAPX1 activity

Liu et al. BMC Plant Biology (2019) 19:488 Page 4 of 13

damage than that of the transgenic lines. Total chlorophylland relative water content (RWC) also considerably de-creased in the WT plants compared with the transgeniclines (Fig. 7b and c). The survival rate was used to furtherdemonstrate the drought tolerance of transgenic lines, andresults showed that the two transgenic lines retainedhigher survival rate after treatment (Fig. 7d).

DiscussionThe APX protein contains hemoglobin that possesses ahigh specificity to ascorbic acid. APX also has a higheraffinity with H2O2 and utilizes AsA as a specific electrondonor, which plays a crucial role in modulating or elim-inating H2O2 from cells and maintaining cellular redoxequilibrium [33–35]. Celery, which is rich in AsA, iswidely cultivated worldwide [29]. In the present work,the AgAPX1 gene, was cloned from celery cv. ‘Jinnanshi-qin’. The prediction of AgAPX1 specific structure willhelp us understand its particular function. AgAPX1 ishighly conserved and characterized with a peroxidasedomain. The deduced amino acid sequence of AgAPX1was the highest similar to that of APX from P. sativum.

AgAPX1 is very conservative and shows a high hom-ology to other APXs. Phylogenic analysis indicated thatthe APX protein of the same family is closer in evolutionand the sequence divergence of APX might occur afterthe monocot–dicot split as other structural genes [36].Several reports reported the purification of APX from

plant tissues or E. coli cells to investigate its functions[19, 20, 37]. Previous studies indicated that the purifiedAPX has a molecular mass ranging from 28 kDa to 31kDa [20, 21, 38]. The molecular mass of rAgAPX1 fromcelery was higher, which was approximately 34 kDa asmeasured by SDS-PAGE. The optimal temperature andpH of the rAgAPX1 enzyme were 55 °C and 7.0, respect-ively. By contrast, the optimal cytosolic APX activity wasat 38 °C and pH 6.5 in komatsuna [20]. In liverwort, theAPX enzyme had an optimal temperature of 40 °C andan optimal pH of 6.0 [38]. The differences in physico-chemical properties among different species may becaused by the differences in APX gene sequence.Another reason may be affected by the insertion of His-tag, S-tag, enterokinase, and thrombin sequences in theN-terminus of the rAgAPX1 protein [39, 40].

Fig. 4 Subcellular localization analysis of AgAPX1. a Control vector (pA7-GFP) expressed in onion epidermal cells. b Recombinant vector (AgAPX1-GFP)expressed in onion epidermal cells. c AgAPX1-GFP fusion proteins transiently expressed in Arabidopsis mesophyll protoplasts. GFP: green fluorescentprotein; BF: bright field; chloroplast: chlorophyll auto fluorescence; merged: combined fluorescence from GFP, BF, and chloroplast

Liu et al. BMC Plant Biology (2019) 19:488 Page 5 of 13

The analysis of gene expression is an important strat-egy to understand the molecular mechanisms of stressresponses in higher plants. Previous studies have re-ported that plants generally exhibited increased APX ac-tivity under stress conditions, which is usually correlatedwith increased stress tolerance [41, 42]. Simultaneously,the expression of APX genes may be induced to protectagainst oxidative stress [42]. The activity of APX en-zymes is enhanced when subjected to cold and saltstresses [16, 43]. In addition, the expression of the APXgene is upregulated under drought stress in Eleusinecoracana [44]. In the present study, the transcriptionlevel of the AgAPX1 gene was markedly upregulatedwhen exposed to PEG-induced dehydration stress in ‘Jin-nanshiqin’ plants, which is consistent with the results ofprevious researches [45, 46]. The result suggests that theAgAPX1 gene involved in the response of celery todrought stress. The response of the AgAPX1 gene toadversity may be attributed to the conservation of APXsequences among different species. In addition, most ofthe differentially expressed genes related to antioxidantenzymes are localized in the mitochondria, chloroplast,

and peroxisome [47, 48]. APXs have several forms of iso-enzymes according to their distribution in plant cells, in-cluding stromal (sAPX), thylakoid membrane (tAPX) inthe chloroplast, microbody APX (mAPX), and cytosolicAPX (cAPX) [17]. Our study indicated that AgAPX1 ismainly localized in the nucleus and membrane. The prob-able functions of APX in the nucleus are to orchestrateROS signaling and regulate related genes expression,thereby regulating plant physiological changes and stressresponse [49, 50].Drought stress is a global problem that limits horticul-

tural productivity. Some defense mechanisms in plantsare aroused under stress to protect themselves from thedamage of oxidative stress [51]. Previous studies also re-ported that the overexpression of APX in plants en-hances the tolerance to both cold and heat stresses [50].Overexpression of the SsAPX gene isolated from Suaedasalsa in Arabidopsis protects plants against salt-inducedoxidative stress [52]. AsA is a natural antioxidant andacts directly to neutralize superoxide radicals in plant re-sponse to abiotic stress [24]. The overexpression ofgenes involved in AsA metabolism, such as GMP [36],

Fig. 5 PCR amplification, GUS stain identification, and functional analysis of Arabidopsis transformed with a AgAPX1 gene. a PCR amplification ofAgAPX1 from the cDNA of transgenic Arabidopsis plants. b Histochemical GUS assays of control line and transgenic lines. c Ascorbate levels of 4-weeks-old transgenic Arabidopsis and WT plants. d The difference of antioxidant capacity between 4-weeks-old transgenic Arabidopsis and WTplants. WT, non-transgenic Arabidopsis (negative control); AgAPX1–4 and AgAPX1–16, transgenic Arabidopsis plants. Error bars represent standarddeviation among three independent replicates. Data are expressed as the means ± standard deviation (SD) of three replicates. Different lettersrepresent significant difference at 0.05 level

Liu et al. BMC Plant Biology (2019) 19:488 Page 6 of 13

GGP [53], and DHAR [54], provides an elevated resist-ance to abiotic stresses by regulating the generation of AsAin plants. We found that overexpressing AgAPX1 in Arabi-dopsis showed increased AsA content significantly. Underdrought stress, the transgenic Arabidopsis (AgAPX1–4 andAgAPX1–16) lines displayed better growth status than theWT plants. Multiple enzymatic antioxidants play key rolesin scavenging the toxic ROS and protecting the plantsfrom oxidative damage [55]. The enzyme activity of

APX in transgenic Arabidopsis also increased. AgAPX-16 line showed higher APX activities than AgAPX-4line. The possible reason is that AsA biosynthesis isregulated complexly by many factors aside from APXactivity [29]. Drought stress induced a prominent increasein the activities of some antioxidant enzymes, includingAPX, SOD, and POD in the transgenic Arabidopsis plants.These results suggest that transgenic Arabidopsis plantshosting AgAPX1 had an increased tolerance to drought

Fig. 6 Phenotype, stomata state and activity of antioxidant enzymes in wild-type (WT), transgenic lines (AgAPX1–4 and AgAPX1–16) before andafter drought treatment. a Comparison of growth phenotype among transgenic and non-transgenic Arabidopsis lines. For drought stresstreatments, these plants were treated with 400mM mannitol and photographed after 7 days. b Stomatal aperture measurements of differentArabidopsis lines in response to 24 h of drought. c APX activity assay (after 24 h of treatment). d SOD activity assay (after 24 h of treatment). ePOD activity assay (after 24 h of treatment). f CAT activity assay (after 24 h of treatment)

Liu et al. BMC Plant Biology (2019) 19:488 Page 7 of 13

stress by enhancing AsA accumulation and the activitiesof antioxidant enzymes. AgAPX1–16 line showed a lowerincrement of APX activity than the AgAPX1–4 line inresponse to drought. The difference in APX activity oftransgenic lines may be associated with insertion site,post-transcription, and post-translation of the target gene.The WT plants showed a greater elevation of CAT activitythan the transgenic plants. CAT, a direct H2O2 scavengingenzyme, was used to eliminate excessive H2O2 generatedfrom drought stress in the anti-oxidative defense systemof plants [56].The increased stress tolerance is also probably associ-

ated with the closure of stomata in plants. The key role ofstomatal closure in plant innate immunity has been de-scribed [57]. In higher plant, abiotic stress could inducestomatal closure. CO2 levels are decreased in the leaves,

which promotes oxygenation of RuBP to produce add-itional H2O2 [58, 59]. ROS species such as H2O2 aremainly targeted by APXs [33]. Therefore, the decreasedstomatal apertures and elevated antioxidant enzyme activ-ities of the transgenic Arabidopsis plants under droughtstress are justifiable. Meanwhile, the drought-induced re-duction in net photosynthetic rate, relative water content,and chlorophyll content was much lower in the transgeniclines than in the WT plants. These results agree with pre-vious reports [60, 61]. Less chlorophyll degradation anddecreased lipid peroxidation have been detected in trans-genic plants overexpressing APX gene under several ad-versities in previous studies [62, 63]. We concluded thatthe change in AgAPX1 transcriptional level is an import-ant regulatory mechanism in response to abiotic stress incelery and transgenic Arabidopsis.

Fig. 7 Physiological changes in the leaves of wild-type (WT) and transgenic Arabidopsis lines (AgAPX1–4 and AgAPX1–16) subjected to drought. aNet photosynthetic rate. b Total chlorophyll content. c Relative water content. d Plant survival rate

Liu et al. BMC Plant Biology (2019) 19:488 Page 8 of 13

In addition, gene silencing and knockout are importanttechnical means to verify the function of genes in plants.Gene silencing is usually achieved using RNA interfer-ence (RNAi) [64]. And the CRISPR/Cas9 system hasbeen used routinely in the knockout of genes [65]. How-ever, the two technologies are both performed based onthe transgenic technology and an efficient transform-ation system has not yet been established in celery. Wewill focus on eliminating this limitation to generatestable loss-of-function plants for further functional veri-fication in our future work.

ConclusionA key enzyme of AsA–GSH cycle in celery, AgAPX1,was identified and characterized in this study. Thecurrent results provided important information on thepotential characteristics of the APX enzyme. The re-sponse of AgAPX1 to drought in celery and the heterol-ogous expression of the AgAPX1 gene in Arabidopsisfurther elucidated the important roles of AgAPX1.AgAPX1 transgenic lines may establish a reference forselecting crop plants with high drought tolerance andincrease the confidence and basis of its homologousexpression by transgenic technology to enhance stresstolerance in celery. These findings are important forfarming in areas under drought stress.

MethodsPlant materials and stress treatmentArabidopsis Columbia ecotype (the control Arabidopsis)and celery cv. ‘Jinnanshiqin’ were deposited at the StateKey Laboratory of Crop Genetics and Germplasm En-hancement, Nanjing Agricultural University (32°04′N,118°85′E). The plants were grown in pots containing amixture of organic soil and vermiculite (3:1, v/v) in anartificial climate chamber (a 16-h photoperiod, an illu-mination of 240 μmol m− 2 s− 1, day/night temperaturesof 25/16 °C and a relative humidity of 70%). For dehy-dration stress, two-month-old celery plants were treatedwith 20% PEG 6000. The leaf blades of ‘Jinnanshiqin’were harvested after 1, 2, 4, 8, 12, and 24 h and immedi-ately frozen in liquid nitrogen and stored at -80 °C untilanalysis. The Arabidopsis seeds were sown on MS solidmedium for 7 days, transferred to a commercial pottingsoil mixture, and then placed in a controlled growthchamber at 22 °C with a relative humidity of 70% underthe condition (12 h: 12 h, light: dark). Four-week-old A.thaliana seedlings were irrigated with 400mM mannitolsolution. The samples were collected at 24 h aftertreatment for further analysis.

Gene searching, cloning, and bioinformatics analysisThe amino acid sequences of the DcAPX (Accession No.AKH49594.1), an APX from carrot, was used to blast

against our celery transcriptome database to obtain thegene encoding APX [66, 67]. The open reading frame(ORF) in comp120_c0_seq1 was the most similar to thesequence of DcAPX. This ORF was designated as AgAPX1in this study. The gene was cloned from the cDNA of‘Jinnanshiqin’ celery by PCR amplification. The cloningprimers of AgAPX1 were 5′-ATGGGAAAGTGCTATCCAATTGT-3′ and 5′-TTAGGCCTCAGCAAACCCAAGT-3′. Then, the basic physical and chemical proper-ties of the predicted protein were analyzed using ExPASy(http://expasy.org/tools/). The functional domains of theAgAPX1 protein were analyzed using the SMART pro-gram. The three-dimensional structure modeling wasperformed using CPHmodels 3.2 Server (http://www.cbs.dtu.dk/services/CPHmodels/) online analysis software.The amino acid sequences of APX proteins from differentspecies were aligned in the Clustal X program, and aphylogenetic tree was constructed using MEGA 6.0 [68].

Expression patterns of AgAPX1 geneThe RT-qPCR system was performed on the CFX96Real-Time PCR system (Bio-Rad) with SYBR Premix ExTaq. RNA was extracted using the RNAprep pure plantkit (Tiangen-bio, Beijing, China) in accordance with themanufacturer’s instructions, followed by reverse tran-scription to cDNA. A 15 × diluted cDNA sample wasused for RT-qPCR analysis. The AgActin gene was usedas an internal standard for normalization [69]. The pri-mer pairs of the AgAPX1 gene for RT-qPCR were 5′-GCCGCTTGCCTGATGCTACTT-3′ and 5′-CCTTCAAACCCAGAACGCTCCTT-3′. The relative expressiondata were calculated using the 2−ΔΔCt method [70]. Eachsample was performed with three biological replicates.

Expression of AgAPX1 in E. coliThe AgAPX1 gene was cloned into the vector pET-30a(+) between the Bam HI and Sac I sites. The recom-binant forward primer and reverse primer were 5′-GCCATGGCTGATATCGGATCCATGGGAAAGTGCTATCCAATTGT-3′ and 5′-GCAAGCTTGTCGACGGAGCTCTTAGGCCTCAGCAAACCCAAGT-3′, respectively.The amplified vector was then transformed into E. coliDH5α and confirmed by DNA sequencing. E. coliBL21(DE3) cells (TransGen, Beijing, China) were usedfor the expression of pET-30a (+)-AgAPX1. The trans-formed bacterial cells were grown in 50 mL of LBmedium containing 50 mg·L− 1 kanamycin at 37 °C forabout 4 h with 230 rpm shaking until the OD600 valuereached 0.4–0.6. The recombinant protein was induced byadding isopropyl β-D-1-thiogalactopyranoside (IPTG) to afinal concentration of 1.0 mM at 18 °C, and the culturewas continued to shake (220 rpm) for over 12 h.

Liu et al. BMC Plant Biology (2019) 19:488 Page 9 of 13

http://expasy.org/tools/http://www.cbs.dtu.dk/services/CPHmodels/http://www.cbs.dtu.dk/services/CPHmodels/

Purification of recombinant protein AgAPX1All steps of the recombinant protein purification werecarried out at 4 °C. The cells were harvested by centrifu-gation; suspended in 4mL of lysis buffer (pH = 7.5) con-taining 50 mM NaH2PO4, 300 mM NaCl, 10% glycerol,10 mM β-mercaptoethanol and 10mM imidazole; anddisrupted by sonication for 20 min on ice. Cell hom-ogenate after sonication was centrifuged at 12,000 rpmfor 10 min at 4 °C, and the supernatant was filtered using0.22-μm microfiltration membranes. The mixture wasloaded on a column containing Ni-NTA-agarose resin(1.5 mL bed volume) (Qiagen, Hilden, Germany) equili-brated with equilibrium buffer (50mM NaH2PO4, 300mMNaCl, 10% glycerol and 10mM imidazole, pH= 7.5).Washing with a buffer (pH = 7.5) containing 50 mMNaH2PO4, 300 mM NaCl, 10% glycerol, 10 mM β-mercaptoethanol, and 50 mM imidazole was performedmore than six times to elute other proteins. Then, theHis-tagged AgAPX1 was dissolved by elution buffer(50 mM NaH2PO4, 300 mM NaCl, 10% glycerol, 10 mMβ-mercaptoethanol and 10 mM imidazole, pH = 7.5)and subsequently washed with equilibrium buffer to re-equilibrate the column. The purified enzyme was collectedas previously described and then stored at -20 °C for fur-ther analysis. The purity of the active fractions was testedusing 12% (w/v) SDS-PAGE as described by Laemmli [71]and then stained with Coomassie Brilliant Blue.

AgAPX1 activity assayAPX activity was measured as previously described byNakano & Asada (1987) [72] with some modifications.Briefly, the reaction mixture contained 50 mM PBS(pH = 7.0) with 0.1 mM EDTA-Na2, 0.5 mM AsA and0.1 mM H2O2, and about 50 μg purified rAgAPX1 in afinal volume of 300 μL. The reaction was initiated withthe addition of H2O2. Enzyme activity was determinedby measuring the decrease in absorbance at 290 nm(Ɛ290/2.8 mM

− 1 cm− 1) due to AsA oxidation over a fewminutes. Protein concentration was determined withCoomassie Brilliant Blue G-250 according to the methodof Bradford using bovine serum albumin (BSA) as thestandard.

Effect of pH and temperature on AgAPX1 activityThe optimum pH for enzyme activity was determinedin 50 mM citric acid–Na2HPO4 (pH = 4.0–5.0), 50 mMNaH2PO4–Na2HPO4 (pH = 6.0–8.0), and 50 mM gly-cine–NaOH (pH = 9.2). The effect of temperature(20 °C–90 °C) on enzyme activity was determined atpH 7.0 in 50 mM PBS for 5 min.

Subcellular localizationThe full-length sequence of the AgAPX1 gene withoutthe terminator was amplified and then inserted into the

GFP-fusion expression vector pA7. The recombinant vec-tor (AgAPX1-GFP) and empty vector (pA7-GFP) weretransferred into onion cells with gold powder via a biolis-tic procedure by using a helium driven particle accelerator(PDS-1000, Bio-Rad). The onion epidermal cells werespread on the MS solid medium for over 16 h in a dark in-cubator at 25 °C. Moreover, the recombinant vector 35S:AgAPX1-GFP was transformed into Arabidopsis meso-phyll protoplasts to identify the real location of AgAPX1since some APXs are located in chloroplast and onionepidermal cells lack chloroplasts. The isolation of Ara-bidopsis mesophyll protoplasts and transient geneexpression were conducted according to PEG-mediatedtransformation method described by Yoo et al. [73]with some modifications. The transformed protoplastsalso were incubated in a dark incubator at 25 °C forover 16 h. The fluorescence signal of GFP fusion pro-teins and red chloroplast auto fluorescence were ob-served using a LSM780 confocal microscopy imagingsystem (Zeiss Co., Oberkochen, Germany).

Overexpression vector construct and ArabidopsistransformationThe full-length ORF of AgAPX1 was amplified usingthe specific primers (forward: 5′-TTTACAATTACCATGGGATCCATGGGAAAGTGCTATCCAATTGT-3′ andreverse:5′-ACCGATGATACGAACGAGCTCTTAGGCCTCAGCAAACCCAAGT-3′) and subcloned into the vectorpCAMBIA-1301. The construct was verified by PCR andsequencing. The recombinant vector (CaMV 35S:AgAPX1)containing a 35S: GUS reporter gene sequence was intro-duced in the Agrobacterium tumefaciens strain GV3101 viaelectroporation. Transformation of WT plants was per-formed using the floral-dip method as previously described[74]. The transgenic lines of Arabidopsis were screened ona MS medium containing hygromycin (40mg/L) and thenplanted. The seeds were then harvested. To further verifythe presence of AgAPX1 in the transgenic Arabidopsis,GUS staining and PCR amplification were conducted. Theindependent transgenic Arabidopsis lines obtained byscreening were used in the next experiment.

Determination of AsA contentAsA content was determined in accordance with themethod described by Duan with slight modifications[75]. Briefly, about 0.3 g fresh leaves from transgenic andWT Arabidopsis plants were ground with 2 mL 1% (w/v)pre-cooling oxalic acid. The homogenate was centri-fuged at 12,000 rpm for 10min at 4 °C and the super-natant was filtered through a 0.45 μm membrane syringefilter. Sample analysis was performed using the Agilent1200 HPLC system for assays of AsA levels at a wave-length of 245 nm. The mobile phase consisted of 0.1%

Liu et al. BMC Plant Biology (2019) 19:488 Page 10 of 13

(v/v) acetic acid at a flow rate of 1 ml/min. AsA contentwas expressed as mg/100 g (fresh weight, Fw).

Antioxidant capacity analysisThe antioxidant capability of the samples was estimatedby the FRAP (ferric reducing antioxidant potential)method. It was performed using the T-AOC Assay Kit(S0116, Beyotime, Shanghai, China) in accordance withthe manufacturer’s instructions. Approximately 0.2 gfresh leaves of Arabidopsis plants were homogenatedwith 1 mL PBS (pH = 7.0), and the absorbance was mon-itored using a spectrophotometer at 593 nm. The resultof antioxidant activity was recorded as mM Fe2+ per gfresh weight (mM Fe2+/g Fw).

Analysis of stomatal characteristicsImages of the stomata were obtained using the abaxialepidermis of leaves detached from 4-week-old Arabidop-sis plants before and after 24 h of drought stress. Theobservation and statistics of stomatal lengths and aper-tures were accomplished randomly using electron mi-croscopy and CellSens image software.

Crude enzyme extract and enzymatic activitymeasurementThe crude enzyme was extracted from 4-week-old leavesof both transgenic and non-transgenic Arabidopsis plantsafter 400mM mannitol for 24 h. For SOD and POD activ-ity determination, about 0.2 g fresh samples were extractedwith 50mM PBS (pH 7.8) and detected using Wang’smethod [76]. For the determination of APX and CAT ac-tivities, fresh Arabidopsis leaves were homogenated with50mM PBS (pH 7.0, containing 0.1mM EDTA). The testof crude APX activity was performed the same as that de-scribed above. CAT activity assay was performed followingthe modified method of Sahu et al. [77]. Each experimentwas performed with three independent replicates.

Measurements of photosynthetic rate, total chlorophyllcontent, RWC, and plant survival ratePn was measured using the Li-6400 portable photo-synthesis system. Total chlorophyll content was deter-mined following the method of Lichtenthaler andWellburn [78]. RWC was measured as previouslydescribed [79]. It was calculated according to the for-mula as follow: RWC = [Fw-Dw]/[Tw-Dw]X100, whereFw is fresh weight of leaf, and Tw is rehydrated weightof leaf after incubating in water for 8 h, and then dry-ing them in an oven at 80 °C until constant weight isrecorded as Dw. In addition, the Arabidopsis plantswere treated by withholding watering for 10 d, andthen the survival rate was recorded after rewateringfor 3 d after recovery. All the above experiments werecarried out using 4-week-old leaves.

Statistical analysisAll values reported in this study were the means of threeindependent replicate measurements, unless otherwisestated. The data were expressed as mean ± standarddeviation (SD) of three replicates. Statistical significanceof the differences was analyzed by SPSS 20.0 software(SPSS Inc., Chicago, IL, USA) using Duncan’s multiple-range test with a significance level of 0.05 (P < 0.05).

Supplementary informationSupplementary information accompanies this paper at https://doi.org/10.1186/s12870-019-2095-1.

Additional file 1 Nucleotide acid and deduced amino acid sequence ofAgAPX1 from celery.

Additional file 2 SDS-PAGE analysis of the purified AgAPX1 fromexpression in E. coli. Lane1, standard markers; Lane2, total soluble proteinfrom BL21(DE3) cells containing the AgAPX1 plasmid without induction;Lane3, total soluble protein from BL21(DE3) cells containing the AgAPX1plasmid with IPTG; Lane 4, purified AgAPX1. The arrow indicates theAgAPX1.

Additional file 3 Ascorbate content in transgenic Arabidopsis and wild-type (WT) leaves detected by HPLC. a WT plants; b AgAPX1–4 transgenicline; c AgAPX1–16 transgenic line.

AbbreviationsAPX: Ascorbate peroxidase; AsA: Ascorbate; BSA: Bovine serum albumin;CAT: Catalase; DHA: Dehydroascorbate; EDTA: Ethylenediaminetetraaceticacid; Fw: Fresh weight; GUS: β-glucuronidase stain; IPTG: Isopropyl β-D-1-thiogalactopyranoside; MDA: Monodehydroascorbate; PBS: Phosphate buffersolution; Pn: Net photosynthetic rate; POD: Peroxidase;rAgAPX1: recombinant AgAPX1 protein; ROS: Reactive oxygen species;rpm: revolution per minute; RWC: Relative water content; SOD: Superoxidedismutase

AcknowledgementsNot applicable.

Authors’ contributionsASX and JXL initiated and designed the research, JXL, AQD, HL, and QQYperformed the experiments; JXL, KF, and ZSX analyzed the data; ASXcontributed reagents/materials/analysis tools; JXL wrote the paper; ASX andKF revised the paper. All authors read and approved the final manuscript.

FundingThis study was financially supported by Jiangsu Agricultural Science andTechnology Innovation Fund [CX (18)2007], New Century Excellent Talents inUniversity (NCET-11-0670), National Natural Science Foundation of China(31272175), Priority Academic Program Development of Jiangsu HigherEducation Institutions Project (PAPD).

Availability of data and materialsThe data sets supporting the conclusions of this article are included withinthe article and its additional files.Arabidopsis Columbia ecotype (the control Arabidopsis) and celery‘Jinnanshiqin’ were deposited at the State Key Laboratory of Crop Geneticsand Germplasm Enhancement, Nanjing Agricultural University.

Ethics approval and consent to participateNot applicable.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Liu et al. BMC Plant Biology (2019) 19:488 Page 11 of 13

https://doi.org/10.1186/s12870-019-2095-1https://doi.org/10.1186/s12870-019-2095-1

Received: 30 March 2019 Accepted: 23 October 2019

References1. Gill SS, Tuteja N. Reactive oxygen species and antioxidant machinery in

abiotic stress tolerance in crop plants. Plant Physiol Biochem. 2010;48(12):909–30.

2. Petrov VD, Van Breusegem F. Hydrogen peroxide-a central hub forinformation flow in plant cells. AoB Plants. 2012;2012(1):pls014.

3. Song XS, Hu WH, Mao WH, Ogweno JO, Zhou YH, Yu JQ. Response ofascorbate peroxidase isoenzymes and ascorbate regeneration system toabiotic stresses in Cucumis sativus L. Plant Physiol Biochem. 2005;43(12):1082–8.

4. Yadav P. Partial purification and characterization of ascorbate peroxidasefrom ripening ber (Ziziphus mauritiana L) fruits. Afr J Biotechnol. 2014;13(32):3323–31.

5. Ishikawa T, Shigeoka S. Recent advances in ascorbate biosynthesis and thephysiological significance of ascorbate peroxidase in photosynthesizingorganisms. Biosci Biotechnol Biochem. 2008;72(5):1143–54.

6. Lazzarotto F, Teixeira FK, Rosa SB, Dunand C, Fernandes CL, Fontenele AV,Silveira JA, Verli H, Margis R, Margis-Pinheiro M. Ascorbate peroxidase-related (APx-R) is a new heme-containing protein functionally associatedwith ascorbate peroxidase but evolutionarily divergent. New Phytol. 2011;191(1):234–50.

7. Lobo V, Patil A, Phatak A, Chandra N. Free radicals, antioxidants and functionalfoods: impact on human health. Pharmacogn Rev. 2010;4(8):118–26.

8. Braun HJ, Rajaram S, Ginkel MV. CIMMYT's approach to breeding for wideadaptation. Euphytica. 1996;92(1–2):175–83.

9. Thoen MP, Davila Olivas NH, Kloth KJ, Coolen S, Huang PP, Aarts MG, Bac-Molenaar JA, Bakker J, Bouwmeester HJ, Broekgaarden C, et al. Geneticarchitecture of plant stress resistance: multi-trait genome-wide associationmapping. New Phytol. 2017;213(3):1346–62.

10. Guan Q, Wang Z, Wang X, Takano T, Liu S. A peroxisomal APX fromPuccinellia tenuiflora improves the abiotic stress tolerance of transgenicArabidopsis thaliana through decreasing of H2O2 accumulation. J PlantPhysiol. 2015;175:183–91.

11. Faize M, Burgos L, Faize L, Piqueras A, Nicolas E, Barbaespin G,Clementemoreno MJ, Alcobendas R, Artlip T, Hernandez JA. Involvement ofcytosolic ascorbate peroxidase and cu/Zn-superoxide dismutase forimproved tolerance against drought stress. J Exp Bot. 2011;62(8):2599–613.

12. Chiang CM, Chien HL, Chen LFO, Hsiung TC, Chiang MC, Chen SP, Lin KH.Overexpression of the genes coding ascorbate peroxidase from Brassicacampestris enhances heat tolerance in transgenic Arabidopsis thaliana. BiolPlantarum. 2015;59(2):305–15.

13. Hu Y, Wu Q, Sprague SA, Park J, Oh M, Rajashekar CB, Koiwa H, Nakata PA,Cheng N, Hirschi KD, White FF, Park S. Tomato expressing Arabidopsisglutaredoxin gene AtGRXS17 confers tolerance to chillingstress viamodulating cold responsive components. Hortic Res. 2015;2:15051.

14. Wang LY, Zhang QY, Wang F, Meng X, Meng QW. Ascorbate plays a keyrole in alleviating low temperature-induced oxidative stress in Arabidopsis.Photosynthetica. 2012;50(4):602–12.

15. Mittler R, Zilinskas BA. Purification and characterization of pea cytosolicascorbate peroxidase. Plant Physiol. 1991;97(3):962–8.

16. Mohammadian MA, Largani ZK, Sajedi RH. Quantitative and qualitativecomparison of antioxidant activity in the flavedo tissue of three cultivars ofcitrus fruit under cold stress. Aust J Crop Sci. 2012;6(3):402–6.

17. Shigeoka S, Ishikawa T, Tamoi M, Miyagawa Y, Takeda T, Yabuta Y,Yoshimura K. Regulation and function of ascorbate peroxidase isoenzymes.J Exp Bot. 2002;53(372):1305–19.

18. Chen GX, Asada K. Ascorbate peroxidase in tea leaves: occurrence of twoisozymes and the differences in their enzymatic and molecular properties.Plant Cell Physiol. 1989;30(7):987–98.

19. Lu Z, Takano T, Liu S. Purification and characterization of two ascorbateperoxidases of rice (Oryza sativa L.) expressed in Escherichia coli. BiotechnolLett. 2005;27(1):63–7.

20. Ishikawa T, Takeda T, Shigeoka S. Purification and characterization ofcytosolic ascorbate peroxidase from komatsuna (Brassica rapa). Plant Sci.1996;120(1):11–8.

21. Leonardis SD, Dipierro N, Dipierro S. Purification and characterization of anascorbate peroxidase from potato tuber mitochondria. Plant PhysiolBiochem. 2000;38(10):773–9.

22. Sun MH, Xu Y, Huang JG, Jiang ZS, Shu HR, Wang HS, Zhang SZ. Globalidentification, classification, and expression analysis of MAPKKK genes:functional characterization of MdRaf5 reveals evolution and drought-responsive profile in apple. Sci Rep. 2017;7(1):13511.

23. Athar HUR, Khan A, Ashraf M. Exogenously applied ascorbic acid alleviatessalt-induced oxidative stress in wheat. Environ Exp Bot. 2008;63(3):224–31.

24. Shalata A, Neumann PM. Exogenous ascorbic acid (vitamin C) increasesresistance to salt stress and reduces lipid peroxidation. J Exp Bot. 2001;52(364):2207–11.

25. Yan H, Li Q, Park SC, Wang X, Liu YJ, Zhang YG, Tang W, Kou M, Ma DF.Overexpression of CuZnSOD and APX enhance salt stress tolerance in sweetpotato. Plant Physiol Biochem. 2016;109:20–7.

26. Jiang Q, Xu ZS, Wang F, Li MY, Ma J, Xiong AS. Effects of abiotic stresses onthe expression of Lhcb1 gene and photosynthesis of Oenanthe javanica andApium graveolens. Biol Plantarum. 2014;58(2):256–64.

27. Li MY, Hou XL, Wang F, Tan GF, Xu ZS, Xiong AS. Advances in the researchof celery, an important Apiaceae vegetable crop. Crit Rev Biotechnol. 2018;38(2):172–83.

28. Li MY, Feng W, Qian J, Jing M, Xiong AS. Identification of SSRs anddifferentially expressed genes in two cultivars of celery (Apium graveolens L.)by deep transcriptome sequencing. Hortic Res. 2014;1:10.

29. Huang W, Wang GL, Li H, Wang F, Xu ZS, Xiong AS. Transcriptional profilingof genes involved in ascorbic acid biosynthesis, recycling, and degradationduring three leaf developmental stages in celery. Mol Gen Genomics. 2016;291(6):2131–43.

30. Robert X, Gouet P. Deciphering key features in protein structures with thenew ENDscript server. Nucleic Acids Res. 2014;42(Web Server issue):W320–4.

31. Xing Y, Jia WS, Zhang JH. AtMEK1 mediates stress-induced gene expressionof CAT1 catalase by triggering H2O2 production in Arabidopsis. J Exp Bot.2007;58(11):2969–81.

32. Cantarello C, Volpe V, Azzolin C, Bertea C. Modulation of enzyme activitiesand expression of genes related to primary and secondary metabolism inresponse to UV-B stress in cucumber (Cucumis sativus L.). J Plant Interact.2005;1(3):151–61.

33. Mittler R. Oxidative stress, antioxidants and stress tolerance. Trends Plant Sci.2002;7(9):405–10.

34. Senthil Kumar R, Shen CH, Wu PY, Suresh Kumar S, Hua MS, Yeh KW. Nitricoxide participates in plant flowering repression by ascorbate. Sci Rep. 2016;6:35246.

35. Sofo A, Scopa A, Nuzzaci M, Vitti A. Ascorbate peroxidase and catalaseactivities and their genetic regulation in plants subjected to drought andsalinity stresses. Int J Mol Sci. 2015;16(6):13561–78.

36. Wang HS, Yu C, Zhu ZJ, Yu XC. Overexpression in tobacco of a tomatoGMPase gene improves tolerance to both low and high temperature stressby enhancing antioxidation capacity. Plant Cell Rep. 2011;30(6):1029–40.

37. Sohn SI, Kim JC, Lee KW, Rhee HI, Wang MH. Molecular cloning andexpression of a cDNA encoding cytosolic ascorbate peroxidase fromPimpinella brachycarpa. J Plant Physiol. 2002;159(9):1029–35.

38. Sajitha Rajan S, Murugan K. Purification and kinetic characterization of theliverwort Pallavicinia lyelli (hook.) S. gray. Cytosolic ascorbate peroxidase.Plant Physiol Biochem. 2010;48(9):758–63.

39. Feng K, Xu ZS, Liu JX, Li JW, Wang F, Xiong AS. Isolation, purification, andcharacterization of AgUCGalT1, a galactosyltransferase involved inanthocyanin galactosylation in purple celery (Apium graveolens L.). Planta.2018;247(6):1363–75.

40. Xu ZS, Jing M, Feng W, Ma HY, Wang QX, Xiong AS. Identification andcharacterization of DcUCGalT1, a galactosyltransferase responsible foranthocyanin galactosylation in purple carrot (Daucus carota L.) taproots. SciRep. 2016;6:27356.

41. Allen RD. Dissection of oxidative stress tolerance using transgenic plants.Plant Physiol. 1995;107(4):1049–54.

42. Yoshimura K, Yabuta Y, Ishikawa T, Shigeoka S. Expression of spinachascorbate peroxidase isoenzymes in response to oxidative stresses. PlantPhysiol. 2000;123(1):223–34.

43. Mishra S, Bandhu AD. Effect of short-term exposure to NaCl onphotochemical activity and antioxidant enzymes in Bruguiera parviflora, anon-secretor mangrove. Acta Physiol Plant. 2004;26(3):317–26.

44. Bhatt D, Saxena SC, Jain S, Dobriyal AK, Majee M, Arora S. Cloning,expression and functional validation of drought inducible ascorbateperoxidase (Ec-apx1) from Eleusine coracana. Mol Biol Rep. 2013;40(2):1155–65.

Liu et al. BMC Plant Biology (2019) 19:488 Page 12 of 13

45. Akbudak MA, Filiz E, Vatansever R, Kontbay K. Genome-wide identificationand expression profiling of ascorbate peroxidase (APX) and glutathioneperoxidase (GPX) genes under drought stress in sorghum (Sorghum bicolorL.). J Plant Growth Regul. 2018;37(3):925–36.

46. Akcay UC, Ercan O, Kavas M, Yildiz L, Yilmaz C, Oktem HA, Yucel M.Drought-induced oxidative damage and antioxidant responses in peanut(Arachis hypogaea L.) seedlings. Plant Growth Regul. 2010;61(1):21–8.

47. Palma JM, Jiménez A, Sandalio LM, Corpas FJ, Lundqvist M, Gómez M,Sevilla F, del Río LA. Antioxidative enzymes from chloroplasts, mitochondria,and peroxisomes during leaf senescence of nodulated pea plants. J Exp Bot.2006;57(8):1747–58.

48. Palma JM, Sevilla F, Jiménez A, Río LAD, Corpas FJ, Morales PÁD, CamejoDM. Physiology of pepper fruits and the metabolism of antioxidants:chloroplasts, mitocondria and peroxisomes. Ann Bot. 2015;116(4):627–36.

49. Caverzan A, Passaia G, Rosa SB, Ribeiro CW, Lazzarotto F, Margis-Pinheiro M.Plant responses to stresses: role of ascorbate peroxidase in the antioxidantprotection. Genet Mol Biol. 2012;35(4):1011–9.

50. Wang J, Wu B, Yin H, Fan Z, Li X, Sui N, He L, Li J. Overexpression of CaAPXinduces orchestrated reactive oxygen scavenging and enhances cold andheat tolerances in tobacco. Biomed Res Int. 2017;2017:4049534.

51. Azooz MM, Ismail AM, Elhamd MFA. Growth, lipid peroxidation andantioxidant enzyme activities as a selection criterion for the salt tolerance ofmaize cultivars grown under salinity stress. Int J Agric Biol. 2009;11(1):572–7.

52. Li K, Pang CH, Ding F, Sui N, Feng ZT, Wang BS. Overexpression of Suaedasalsa stroma ascorbate peroxidase in Arabidopsis chloroplasts enhances salttolerance of plants. S Afr J Bot. 2012;78(1–2):235–45.

53. Zhang GY, Liu RR, Zhang CQ, Tang KX, Sun MF, Yan GH, Liu QQ.Manipulation of the rice L -galactose pathway: evaluation of the effects oftransgene overexpression on ascorbate accumulation and abiotic stresstolerance. PLoS One. 2015;10(5):e0125870.

54. Li QZ, Li YS, Li CH, Yu XC. Enhanced ascorbic acid accumulation throughoverexpression of dehydroascorbate reductase confers tolerance to methylviologen and salt stresses in tomato. Czech J Genet Plant. 2012;48(2):74–86.

55. Noctor G, Foyer CH. ASCORBATE AND GLUTATHIONE: Keeping activeoxygen under control. Annu Rev Plant Physiol Plant Mol Biol. 1998;49(1):249–79.

56. Selote DS, Khanna-Chopra R. Drought acclimation confers oxidative stresstolerance by inducing co-ordinated antioxidant defense at cellular andsubcellular level in leaves of wheat seedlings. Physiol Planta. 2006;127(3):494–506.

57. Melotto M, Underwood W, Koczan J, Nomura K, He SY. Plant stomata functionin innate immunity against bacterial invasion. Cell. 2006;126(5):969–80.

58. Huang W, Zhang SB, Hu H. Sun leaves up-regulate the photorespiratorypathway to maintain a high rate of CO2 assimilation in tobacco. Front PlantSci. 2014;5:688.

59. Zhou Y, Huang L, Zhang Y, Shi K, Yu J, Nogués S. Chill-induced decrease incapacity of RuBP carboxylation and associated H2O2 accumulation incucumber leaves are alleviated by grafting onto figleaf gourd. Ann Bot.2007;100(4):839–48.

60. Gu JF, Qiu M, Yang JC. Enhanced tolerance to drought in transgenic riceplants overexpressing C4 photosynthesis enzymes. Crop J. 2013;1(2):105–14.

61. Deng BL, Dong HS. Ectopic expression of riboflavin-binding protein geneTsRfBP paradoxically enhances both plant growth and drought tolerance intransgenic Arabidopsis thaliana. J Plant Growth Regul. 2013;32(1):170–81.

62. Lee DH, Oh DC, Oh YS, Malinverni JC, Kukor JJ, Kahng HY. Cloning andcharacterization of monofunctional catalase from photosynthetic bacteriumRhodospirillum rubrum S1. J Microbiol Biotechnol. 2007;17(9):1460–8.

63. Murgia I, Tarantino D, Vannini C, Bracale M, Carravieri S, Soave C. Arabidopsisthaliana plants overexpressing thylakoidal ascorbate peroxidase showincreased resistance to Paraquat-induced photooxidative stress and to nitricoxide-induced cell death. Plant J. 2004;38(6):940–53.

64. Tang W, Luo XY, Sanmuels V. Gene silencing: double-stranded RNAmediated mRNA degradation and gene inactivation. Cell Res. 2001;11(3):181–6.

65. Xu ZS, Feng K, Xiong AS. CRISPR/Cas9-mediated multiply targetedmutagenesis in orange and purple carrot plants. Mol Biotechnol. 2019;61(3):191–9.

66. Jia XL, Wang GL, Xiong F, Yu XR, Xu ZS, Wang F, Xiong AS. De novoassembly, transcriptome characterization, lignin accumulation, and anatomiccharacteristics: novel insights into lignin biosynthesis during celery leafdevelopment. Sci Rep. 2015;5:8259.

67. Feng K, Hou XL, Li MY, Jiang Q, Xu ZS, Liu JX, Xiong AS. CeleryDB: agenomic database for celery. Database. 2018;2018:bay070.

68. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: molecularevolutionary genetics analysis version 6.0. Mol Biol Evol. 2013;30(12):2725–9.

69. Li MY, Feng W, Qian J, Wang GL, Chang T, Xiong AS. Validation andcomparison of reference genes for qPCR normalization of celery (Apiumgraveolens) at different development stages. Front Plant Sci. 2016;7:313.

70. Pfaffl MW. A new mathematical model for relative quantification in real-timeRT-PCR. Nucleic Acids Res. 2001;29(9):e45.

71. Laemmli UK. Cleavage of structural proteins during the assembly of thehead of bacteriophage T4. Nature. 1970;227(5259):680–5.

72. Nakano Y, Asada K. Purification of ascorbate peroxidase in spinachchloroplasts; its inactivation in ascorbate-depleted medium and reactivationby monodehydroascorbate radical. Plant Cell Physiol. 1987;28(1):131–40.

73. Yoo SD, Cho YH, Sheen J. Arabidopsis mesophyll protoplasts: a versatile cellsystem for transient gene expression analysis. Nat Protoc. 2007;2(7):1565–72.

74. Zhang X, Henriques R, Lin SS, Niu QW, Chua NH. Agrobacterium-mediatedtransformation of Arabidopsis thaliana using the floral dip method. NatProtoc. 2006;1(2):641–6.

75. Li H, Liu ZW, Wu ZJ, Wang YX, Teng RM, Zhuang J. Differentially expressedprotein and gene analysis revealed the effects of temperature on changesin ascorbic acid metabolism in harvested tea leaves. Hortic Res. 2018;5:65.

76. Wang Y, Gao C, Liang Y, Yang C, Liu G. A novel bZIP gene from Tamarixhispida mediates physiological responses to salt stress in tobacco plants. JPlant Physiol. 2010;167(3):222–30.

77. Sahu GK, Sabat SC. Changes in growth, pigment content and antioxidantsin the root and leaf tissues of wheat plants under the influence ofexogenous salicylic acid. Braz J Plant Physiol. 2011;23(3):209–18.

78. Lichtenthaler HK, Wellburn AR. Determinations of total carotenoids andcholorphylls a and b of leaf extracts in different solvents. Biochem SocTrans. 1983;11(5):591–2.

79. Griffiths A, Bray EA. Shoot induction of ABA-requiring genes in response tosoil drying. J Exp Bot. 1996;47(10):1525–31.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Liu et al. BMC Plant Biology (2019) 19:488 Page 13 of 13

AbstractBackgroundResultsConclusions

BackgroundResultsNucleotide sequence and deduced amino acid sequence of AgAPX1Expression profiles of AgAPX1 under PEG 6000 treatment in celeryExpression of AgAPX1 in Escherichia coli and purification of recombinant AgAPX1Characterization of rAgAPX1Subcellular localization of the AgAPX1 proteinHeterologous expression of AgAPX1 in Arabidopsis increased the AsA content and total antioxidant capacityOverexpression of AgAPX1 in Arabidopsis positively regulates drought tolerance by regulating the stomata apertureAnalysis of antioxidant enzyme activities in transgenic lines before and after treatmentPhysiological changes in Arabidopsis leaves exposed to drought stress

DiscussionConclusionMethodsPlant materials and stress treatmentGene searching, cloning, and bioinformatics analysisExpression patterns of AgAPX1 geneExpression of AgAPX1 in E. coliPurification of recombinant protein AgAPX1AgAPX1 activity assayEffect of pH and temperature on AgAPX1 activitySubcellular localizationOverexpression vector construct and Arabidopsis transformationDetermination of AsA contentAntioxidant capacity analysisAnalysis of stomatal characteristicsCrude enzyme extract and enzymatic activity measurementMeasurements of photosynthetic rate, total chlorophyll content, RWC, and plant survival rateStatistical analysis

Supplementary informationAbbreviationsAcknowledgementsAuthors’ contributionsFundingAvailability of data and materialsEthics approval and consent to participateConsent for publicationCompeting interestsReferencesPublisher’s Note